Colorado mountains
From Long-Term Data to Understanding: Toward a Predictive Ecology
2015 LTER ASM Estes Park, CO - August 30 - September 2, 2015

Luquilo: Discoveries Over the Last 27-Years of Tropical Forest Dynamics – Spatial Variability in Tree Species Distributions

Printer-friendly versionPrinter-friendly version

Poster Number: 
Presenter/Primary Author: 
James Aaron Hogan
Jess K. Zimmerman
Aaron Sheils

One main objective of LTER is to document long-term ecological processes at various spatial scales.  The Luquillo (LUQ) site was added to the network in 1988 and remains the only tropical site.  The 16-Ha permanent “Big Grid” was established in 1989 following Hurricane Hugo, which severely damaged the forest, and has been censused 5 times since at 5-year intervals.  The monitoring of succession after hurricane disturbance has revealed compositional variability in the tree community spatially.  Confirmed by historical areal photographs, signatures of past land-use related to previous settlement and logging practices have confounded successionial dynamics, creating communities of unique composition and novel succesional trajectories.  Variance portioning of the tree community with respect to the environment revealed space (e.g. density dependent factors and the spatial autocorrelation of the tree community) explained the majority (30%) of the variance in vegetation community composition (all stems ³ 1 cm dbh) in the most recent tree census (2011).  The variance explained by space was large, when compared to environmental influences (soil resources and topography), which only explained a small fraction of variance in tree community composition (< 3% respectively).

The additional investigation of succession and spatial variability within the vegetation community of LUQ via an experimental hurricane simulation; the Canopy-Trimming Experiment, was established as part of the LUQ LTER in 2002.  The experiment consists of 3 replicates in a 2 x 2 factorial design consisting of four treatments aimed to mimic and identify the effects of two key components of hurricane disturbance: canopy opening and debris deposition to the forest floor.  Initially, the four treatments implemented were trim + debris, trim + no debris, no trim + debris, and no trim + no debris.  Two years of pre-treatment monitoring preceded the cutting of trees within trim treatments, followed by monitoring as trim plots recovered from the simulated hurricane.  Canopy openness due to trimming increased the recruitment of small stems, mainly of understory pioneer species, and later mortality.  Deposition of debris to the forest floor had only small positive effects on basal area increase in + debris treatments.  Successional responses across replicates varied due to spatial differences in species composition, providing weak evidence for the Intermediate Disturbance Hypothesis. 

We affirm the spatial heterogeneity of species-rich Caribbean forest, despite the relatively low levels of species richness of LUQ forests when compared to continental tropical forests, to conclude that large-scale, long-term monitoring is necessary to understanding succesional dynamics.   

Student Poster Competition: